[ARM] 3439/2: xsc3: add I/O coherency support

Patch from Lennert Buytenhek

This patch adds support for the I/O coherent cache available on the
xsc3.  The approach is to provide a simple API to determine whether the
chipset supports coherency by calling arch_is_coherent() and then
setting the appropriate system memory PTE and PMD bits.  In addition,
we call this API on dma_alloc_coherent() and dma_map_single() calls.
A generic version exists that will compile out all the coherency-related
code that is not needed on the majority of ARM systems.

Note that we do not check for coherency in the dma_alloc_writecombine()
function as that still requires a special PTE setting.  We also don't
touch dma_mmap_coherent() as that is a special ARM-only API that is by
definition only used on non-coherent system.

Signed-off-by: Deepak Saxena <dsaxena@plexity.net>
Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

3 files changed, 29 insertions, 1 deletions

diff --git a/arch/arm/mm/consistent.c b/arch/arm/mm/consistent.c
index 8a1bfcd5008..50e6b6bfb2e 100644
--- a/arch/arm/mm/consistent.c
+++ b/arch/arm/mm/consistent.c
@@ -18,6 +18,7 @@
 #include <linux/device.h>
 #include <linux/dma-mapping.h>
 
+#include <asm/memory.h>
 #include <asm/cacheflush.h>
 #include <asm/tlbflush.h>
 #include <asm/sizes.h>
@@ -272,6 +273,17 @@ __dma_alloc(struct device *dev, size_t size, dma_addr_t *handle, gfp_t gfp,
 void *
 dma_alloc_coherent(struct device *dev, size_t size, dma_addr_t *handle, gfp_t gfp)
 {
+        if (arch_is_coherent()) {
+                void *virt;
+
+                virt = kmalloc(size, gfp);
+                if (!virt)
+                        return NULL;
+                *handle =  virt_to_dma(dev, virt);
+
+                return virt;
+        }
+
         return __dma_alloc(dev, size, handle, gfp,
                            pgprot_noncached(pgprot_kernel));
 }
@@ -350,6 +362,11 @@ void dma_free_coherent(struct device *dev, size_t size, void *cpu_addr, dma_addr
 
         WARN_ON(irqs_disabled());
 
+        if (arch_is_coherent()) {
+                kfree(cpu_addr);
+                return;
+        }
+
         size = PAGE_ALIGN(size);
 
         spin_lock_irqsave(&consistent_lock, flags);
diff --git a/arch/arm/mm/mm-armv.c b/arch/arm/mm/mm-armv.c
index 5e5d05bcad5..f14b2d0f369 100644
--- a/arch/arm/mm/mm-armv.c
+++ b/arch/arm/mm/mm-armv.c
@@ -389,6 +389,17 @@ void __init build_mem_type_table(void)
         kern_pgprot = user_pgprot = cp->pte;
 
         /*
+         * Enable CPU-specific coherency if supported.
+         * (Only available on XSC3 at the moment.)
+         */
+        if (arch_is_coherent()) {
+                if (cpu_is_xsc3()) {
+                        mem_types[MT_MEMORY].prot_sect |= PMD_SECT_S;
+                        mem_types[MT_MEMORY].prot_pte |= L_PTE_COHERENT;
+                }
+        }
+
+        /*
          * ARMv6 and above have extended page tables.
          */
         if (cpu_arch >= CPU_ARCH_ARMv6 && (cr & CR_XP)) {
diff --git a/arch/arm/mm/proc-xsc3.S b/arch/arm/mm/proc-xsc3.S
index b9dfce57c27..80873b36c3f 100644
--- a/arch/arm/mm/proc-xsc3.S
+++ b/arch/arm/mm/proc-xsc3.S
@@ -371,7 +371,7 @@ ENTRY(cpu_xsc3_switch_mm)
 ENTRY(cpu_xsc3_set_pte)
         str     r1, [r0], #-2048                @ linux version
 
-        bic     r2, r1, #0xff0
+        bic     r2, r1, #0xdf0                  @ Keep C, B, coherency bits
         orr     r2, r2, #PTE_TYPE_EXT           @ extended page
 
         eor     r3, r1, #L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_WRITE | L_PTE_DIRTY